Functional Customization of Two-Dimensional Materials for Photocatalytic Activation and Conversion of Inert Small Molecules in Air

Air has the advantage of abundance and easy availability, so it is suitable to be used as synthetic raw materials and energy source. However, the triggering of inert small molecules in air, like O 2 , N 2 , CO 2 , is a kinetically complex and energetically challenging multistep reaction. Photocatalysis brings hope for this challenge, but obstacles still remain in many aspects. Here, aiming at the key difficulties of the photocatalytic activation and conversion of these three inert small molecules, i.e. regulating electronic structure, active sites, charge carrier separation and mobility, and reaction energy barrier, we propose the concept of functional customization strategy of ultrathin two-dimensional materials for achieving more efficient activation and better performance, including thickness control, vacancy engineering, doping operation, single-atom site fabrication, and composite construction. The in-depth understanding of the functional customization will provide more profound guidance for designing photocatalysts which specialize in activating and converting inert small molecules.

molecules to products which are more practical and environment-friendly. Until now, through photocatalytic conversion route, we can change O 2 into strongly oxidizing species, (5) use N 2 as raw material to generate nitrogen-containing compounds with high industrial value (such as ammonia and nitrate), (6)(7)(8) and produce CO, methane, methanol, formic acid, ethanol, ethylene, aromatics, etc. by transforming CO 2 . (9)(10)(11)(12) In a word, the appropriate utilization of small molecules, mainly O 2 , N 2 and CO 2 , is essential to the daily functioning of human society. However, there still remain many challenges and obstacles in the inert small molecule utilization process.
The dioxygen, on which we rely for survival, is one of the most typical small molecules.
Reactive oxygen species (ROS), obtained by reduction of O 2 , play an important role in many chemical reactions. ROS mainly include superoxide radical (·O 2 -), hydrogen peroxide (H 2 O 2 ), hydroxyl radical (·OH), and singlet oxygen ( 1 O 2 ). (13)(14)(15)(16) These highly reactive oxygen species can be used as green oxidants to oxidize pollutants, biological macromolecules, organic chemicals and other substances, and have important application prospects in pollutant treatment, photodynamic therapy, and selective oxidation reactions. Therefore, O 2 activation has become an indispensable link in the process of utilizing O 2 resources efficiently. However, triplet dioxygen is the ground state of O 2 , which has a special configuration that the most stable form has two unpaired electrons with parallel spins in highest electron occupied molecular orbital, namely degenerate π* anti-bond orbital. Due to the electronic transition rule, the transition from the singlet state to the triplet state is a spin-forbidden process and thus the application of O 2 is restricted. Meanwhile, the dinitrogen is another small molecule which is difficult to be activated due to its strong N≡N triple bond and non-polarity.(17) Both thermodynamic and kinetic factors prevent the reaction from proceeding. (18,19) In industry, the Harbor-Bosch process is conducted under high reaction temperature and pressure, generating huge energy consumption and serious environmental pollution. (20) Thus, it is highly desirable to develop a less energy-consuming, more eco-friendly and efficient nitrogen 1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58 59 60 F o r R e v i e w O n l y 5 fixation method. By utilizing such method, we could make the active sites of the catalysts, which contain empty orbitals and orbital occupied electrons, accept the electrons given by the bonding orbital of N 2 and back-donate to the π* antibonding orbital of N 2 , forming coordination with N 2 to complete chemical adsorption and then achieve the activation of N 2 by increasing the bond length and reducing the bond energy. (21)(22)(23)(24)(25) Furthermore, CO 2 could be converted into various valuable carbon-containing products through different pathways. (26)(27)(28)(29) At present, the difficulties of activating CO 2 mainly lie in low solubility, (30) thermodynamic limit, (31) and kinetic effects. (32,33) Therefore, the functional customization of photocatalysts to activate and convert inert small molecules becomes urgent and inevitable.
To deal with the intractable problems above, appropriate photocatalysts of advantages are supposed to be designed and synthesized. Recently, atomically ultrathin two-dimensional materials have attracted plenty of attention because of the large specific surface area, abundant active sites, short charge carrier migration distances, and other favorable properties, which were suitable for photocatalytic research. (34)(35)(36)(37) More importantly, due to the simplified and manipulable structure, it is convenient to study the photocatalysis mechanism with the aid of ultrathin two-dimensional materials. Therefore, they have been widely used in inert small molecule photocatalytic activation and conversion. (38)(39)(40) In this review, aiming at the key difficulties of the activation and conversion of these three inert small molecules, we have developed a description system similar to the fourth-order tensor to enumerate, generalize, and integrate various photocatalytic research of inert small molecule activation and conversion. In particular, we summarize the recent progress of the functional customization of ultrathin two-dimensional materials in activating and converting the three typical kinds of inert small molecules (O 2 , N 2 , CO 2 ), and we introduce various regulations of two-dimensional materials in detail. In light of the main photocatalytic reaction process (adsorption, activation, dissociation, formation and desorption of intermediates, and F o r R e v i e w O n l y 6 product distribution), we present some opinions about strategies including thickness regulation, vacancy introduction, doping operation, single-atom catalyst construction, and interface engineering, to direct at specific functions, such as electronic structure, active sites, charge carrier separation and mobility, and reaction energy barriers. The research object (inert small molecules), specific modification function, functional customization strategies, and photocatalytic process, as four dimensions, organically combine to form the description system. Finally, we put forward some prospects that may further improve the performance of two-dimensional materials, hoping to promote its application in the field of inert small molecule activation and conversion research.

Photocatalysis Process and Corresponding Performance-enhancing Strategies
Photocatalysis is using light as energy sources to increase the rate of photoreactions by proper catalysts. A typical mechanism of photocatalysis contains five main steps: (1) The photocatalysts harvest light to produce electron-hole pairs. (2) The photogenerated charge carriers transfer to the surface from the bulk (or recombine halfway). (3) Reactants adsorb at the active sites on the surface of photocatalysts. (4) The surface-reached charge carriers trigger redox reactions from adsorbed reactants to products at the corresponding active sites.
(5) Products desorb to restore the active site state. It is worth noting that the band structure plays a crucial role in the catalytic process, because only the light whose energy is larger than the bandgap energy (E g ) between the highest completely filled valence band (VB) and the lowest empty conduction band (CB) could be harvested, and the redox reactions whose redox potential matches the band structure of photocatalysts could be activated. In general, there are four kernel points worthy of consideration, including yield (rate), efficiency, selectivity, and stability of the whole catalytic reactions. To achieving high performance of the four main purposes above, several properties and functions of photocatalysts could be regulated by seeking out appropriate materials and implementing functional customization.  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59 The substantial process of activating inert small molecules is weakening or directly cleaving the existing bonds in the molecules after adsorption. Considering the steps of photocatalysis, it is consequential to think of that there are several key directions which could be taken into account for functional customization, including: (1) Regulating the electronic structure of the photocatalysts to enhance the light harvest and ensure more reductive electrons or oxidative holes to be quickly trapped in the active centers of the photocatalysts.
(2) Introducing more catalytic active sites into the photocatalysts, while two-dimensional materials have exceptional advantages to obtain an extremely high ratio of catalytic active sites on the surface. (3) Searching suitable strategies which can construct a proper surface structure to tighten the adsorption of reactant molecules, because the charge carriers could only transfer to the inert small molecules which are adsorbed onto the photocatalyst surface.
(4) Accelerating the efficient charge exchange between the photocatalysts and adsorbates to lift the activating rate of inert small molecules. (5) Suppressing the recombination of charge carriers to gain higher activating efficiency, making the best use of the energy input. (6) Lowering the overall reaction energy barrier to make the redox reaction more feasible and regulate the reaction pathway to expected products. (7) Facilitating the products to desorb easily and swiftly from the active sites avoiding the poisoning of photocatalysts, to finally complete the conversion process of inert small molecules. With all the factors combined synergically, it could thereby increase to a higher degree of inert small molecule activation, and reduce the reaction activation energy to the more thorough conversion of inert small molecules.
Along with the discovery of graphene, due to the huge potential for manipulating the intrinsic properties in two-dimensional materials without destroying pristine lattices while creating a high exposed proportion of surface atoms, two-dimensional materials construct a suitable model platform for photocatalysis research. It is possible to use the advantages and characteristics to precisely customize each step of the photocatalytic reaction: (1) The energy band structure of two-dimensional nanomaterials is different from the corresponding threedimensional bulk materials, which is the essence of two-dimensional nanomaterials.
Therefore, it is possible to regulate the energy band structure of the photocatalyst to adjust the light absorption properties, and at the same time adjust the corresponding redox ability of the photogenerated carriers. (2) The thickness of two-dimensional nanomaterials is ultrathin, usually at the level of several atomic layers. After the photogenerated carriers are generated from the bulk phase of the material, the distance they need to transfer to the surface of the photocatalyst is very short, which is conducive to photogenerated charge carriers. The highspeed transport could also reduce the probability of photogenerated carriers recombining Through the functional customization of two-dimensional materials, it is very promising to clarify the nature of photocatalytic reactions and improve the performance of inert small molecule activation and conversion. In fact, many approaches have been implemented in twodimensional material to resolve those key difficulties of photocatalysis. Defect energy level caused by introducing vacancies not only alters the electronic structure of the photocatalysts, but also prevents the recombination of carriers. At the same time, the vacancy can cooperate with the nearby low coordinated atoms to promote the transport of electrons to adsorbates and complete efficient activation and conversion process. (41) Besides, the introduction of  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  60  F  o  r  R  e  v  i  e  w  O  n  l  y   9 heteroatoms, namely doping operation, can provide more active sites for inert small molecules to be adsorbed and activated, where constructing heterostructured composites could achieve the same effect and promote charge carriers separating and transferring. (42)(43)(44)(45)(46) In addition, single-atom sites are widely utilized for constructing good adsorbing locations in photocatalysis research. (47) But it is worth noting that single atoms cannot exist independently, while two-dimensional materials are ideal platforms for supporting single atoms to realize the utmost potential of single-atom sites. The mutual benefits between twodimensional materials and single atoms make their composites proper models for adsorbing sites research, helpful for elucidating the catalytic mechanism objectively.(48) In a word, twodimensional materials are appropriate platforms for functional customization to enhance photocatalytic activation and conversion of inert small molecules.

Approaches for Functional Customization of Two-dimensional Materials
Considering the current problems of photocatalytic activation of inert small molecules, we could start from addressing the performance constraints in photocatalysis, by using thickness regulation, vacancy engineering, doping operations, composite construction, and single-atom sites introduction, to enhance the photocatalytic performance in a targeted manner. Then, we will ultimately actualize functional customization through regulating the electronic structure, active sites, charge carrier separation and mobility of ultrathin two-dimensional materials, and lowering the reaction energy barriers of related photocatalytic reactions. And the functional customization of two-dimensional materials photocatalysts will be a promising direction to achieve higher yield, efficiency, selectivity, and stability of inert small molecule conversion process.

Active site construction.
The active sites, which mainly contain the low coordinated steps, edges, terraces, kinks and corner atoms in catalysts, are the places at which all the reactants adsorb and reactions occur.
The reactant adsorption and product desorption, which to a large extent influence the feasibility and rate of the catalytic reactions, are needed to be enhanced for better photocatalytic performance. It is generally accepted that the sites with more dangling bonds possess higher reactive activity, so the sites with lower coordination number inducing more dangling bonds could be regarded as catalytic active sites. The atomic thickness and ultrahigh specific surface area of atomically two-dimensional materials could make the number of their low coordinated sites, namely active sites, comparable with the total number of atoms, leading to the extremely high ratio of active sites for better catalytic efficiency. (67) In addition, the l y 17 more exposed surface atoms in the two-dimensional materials could easily escape from the lattice and hence inevitably result in the formation of defect structure with low coordination number, which creates even more active sites including inaccessible vacancies in threedimensional bulk materials and also strongly affects the electronic structure of catalysts, lowers their surface energy, enhances the adsorption of reactants, and hence endows them with better stability.(68) Therefore, by functional customization, the two-dimensional materials could serve as an ideal model to investigate the atomic level interplay between active sites and catalytic activity, namely the active sites-catalytic activity relationship.
Thickness control alone could increase the active sites to a certain degree. Liang  69) The CO 2 adsorption isotherms and UV-visible diffuse reflectance spectra revealed that Bi 2 WO 6 atomic layers had 3 times higher CO 2 adsorption capacity relative to bulk Bi 2 WO 6 . When the thickness is down to a single-unit-cell, the ultralarge surface area for the single-unit-cell Bi 2 WO 6 layers with sufficient active sites favored the enhancement of CO 2 adsorption capacity, thus providing the prerequisite to participate in the following reactions of CO 2 activation and conversion.
In addition, due to the diversity of CO 2 reduction products, selectivity is also an important issue in the conversion of CO 2 . However, due to the scaling relationships, which means a positive correlation of the binding abilities between the same active site and multiple intermediates, an increase in reactivity may lead to a decrease in selectivity. In atoms around the vacancies, and thus formed Cu-In bimetal active sites. According to the calculated results shown in Figure 4, it is suggested that the formation of COOH* intermediates was the rate-limiting step for both samples, and the V S -CuIn 5 S 8 single-unit-cell layers had a lower COOH* formation energy than the pristine CuIn 5 S 8 single-unit-cell layers.

Charge carrier separation and mobility enhancement.
The charge carrier mobility from the bulk to the surface of the catalysts directly relates to the rate and efficiency of catalytic reactions, so it requires necessarily improvement for high photocatalytic performance. It is also needful to avoid the charge carrier recombination halfway in the reaction, while fortunately the intrinsic ultrathin two-dimensional architecture could shorten the charge diffusion distance and thus reduce the recombination possibility of electron-hole pairs.(80, 81) However, the relationship between active sites and charge carrier mobility is usually conflicting. The mobility of two-dimensional materials is often hindered by charge scattering mechanism from defects or grain boundaries.(82) Moreover, larger specific surface area and abundant active sites are prone to appear with thinner twodimensional construction, while the overall mobility in this kind of materials is usually lower due to the poor interlayer charge carrier transport.(83) Thus, balancing the benefits between rich active sites and high mobility is highly desirable to achieve efficient photocatalytic properties. So the functional customization, controllable disorder engineering with element incorporation, was proposed to synergistically modulate the charge carrier behavior for obtaining promoted photocatalytic performances.
For example, the ZnIn 2 S 4 atomic-level nanosheets with different vacancies were constructed for photoreduction of CO 2 , where the defect type was determined to be zinc vacancies (V Zn ''), showing a 3.6 times yield improvement than the V Zn ''-poor sample. (84) (Figure 7a, b).

Reaction energy barrier reduction.
The reaction energy barrier, which is obtained by subtracting the free energy of the reactant structure from that of the transition state structure using transition state theory, is essential to the rate and selectivity of catalytic reactions. The word "activation energy" is widely used in the catalysis research as well, which we could treat approximately as the same as reaction energy barrier. The appropriately regulated configuration of two-dimensional materials could reduce the overall reaction energy barrier, and under certain circumstances even convert the endoergic step to an exoergic reaction process, thus changing the reaction pathway to produce the expected substance with high selectivity.  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  Similar to surface vacancy defects, the surface pothole structure can also facilitate the activation of N 2 . The surface pothole can be regarded as the uneven distribution of atoms or the vacancies of multiple atoms, so many dangling bonds will emerge around the pothole, which could form a localized electron enrichment area spontaneously to adsorb N 2 directly and influence the reaction path and thus reaction energy barrier. Liu has successfully synthesized the pothole-rich WO 3 nanosheets by a chemical topological transformation strategy and achieved the direct nitrate synthesis from N 2 under mild conditions.(94) Without any sacrificial agent or precious-metal co-catalysts, the average yield of nitrate was 1.92 mg g -1 h -1 at room temperature. It was found from DFT calculation that the nitrogen fixation process on WO 3 nanosheets followed the photogenerated hole oxidation mechanism. Potholefree WO 3 needed to overcome the considerable reaction energy barriers during the adsorption and activation processes of N 2 (Figure 9a, b), while the pothole-rich WO 3 nanosheets could motivate charge carrier transfer spontaneously due to the pothole structure (reaction state A in  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  For constructing single-atom catalysts, single-atom Pt anchored at the −N 3 sites of stable and ultrathin covalent triazine framework nanosheets (Pt-SACs/CTF) have been successfully synthesized. (95) In general, photocatalytic N 2 reduction reaction takes two associative mechanisms, the distal mechanism and the alternating mechanism, in which nitrogen is consecutively protonated via proton coupled electron process (PCET) without breaking the N≡N until the release of the first NH 3 (Figure 10a). The DFT results indicated the alternating mechanism is slightly favored for the N 2 fixation in the Pt-SACs/CTF catalyst. The calculated Gibbs adsorption energy of NH 3 is 1.4 eV, which was too high for the produced NH 3 desorbing in general case. Affected by regulation, the ammonia is not detected by the gaseous NH 3 but by the NH 4 + ions. Therefore, the last step for N 2 photocatalytic fixation to ammonia leading to a reaction energy barrier decrease from beginning to end of the whole photocatalytic process (Figure 10b, c).
Furthermore, isolated single-atom Co incorporated into Bi 3 O 4 Br atomic layers was successfully prepared, performing light-driven CO 2 reduction with a selective CO formation rate about 4 and 32 times higher than that of atomic layers and bulk counterpart, respectively.(96) The Co single atoms could lower the CO 2 activation energy barrier through stabilizing the COOH* intermediates and tune the rate-determining step from the formation of COOH* to desorption of CO* (Figure 10d). Then, as shown from CO TPD, the single-atom sample showed lower initial desorption temperature and higher total CO detection, indicating that the incorporation of Co single atoms is conducive to CO desorption, and this result is verified by calculating the free energy of CO desorption, demonstrating that this functional customization helps to reduce the reaction energy barrier in the photocatalytic process.
Without loading heteroatom, layer-structured zinc silicate (LZS) nanosheets have been successfully synthesized by a liquid-phase epitaxial growth route for efficient conversion of CO 2 into CO. (97) DFT calculation based on the LZS structure model was used to investigate the possible reaction pathway along with the local configurations. As mentioned above, the formation of adsorbed COOH* is usually a key step in the reduction process of CO 2 to CO.

Summary and Outlook
Atomically ultrathin two-dimensional materials, benefiting from their unique structure, bridge the study of the relationship between structure and activity, which is convenient for guiding the design of photocatalysts and opens a new way for the functional customization of materials with better photocatalytic performance. Therefore, it has great potential in the field of inert small molecule activation and conversion. We here summed up concisely the function available to be regulated as electronic structure, active sites, charge carrier separation and mobility, and reaction energy barriers. We then summarized the progress of the functional customization of two-dimensional materials for activating and converting inert small molecules in air, including the preparation of atomically ultrathin two-dimensional nanosheets based on the principle of thickness control, the introduction of anion or cation vacancies, heteroatom engineering containing doping or loading of heteroatoms to fabricate single-atom catalysts, and composite construction with other materials. It is foreseeable that through exploration and innovation of construction strategies, characterization technology, and dynamics research with the development of theoretical calculations, we could make considerable progress in understanding the mechanism of inert small molecule activation and conversion in two-dimensional materials and regulating their photocatalytic performance.
Finally, how to further improve the yield, efficiency, selectivity, and stability is an urgent subject to be addressed.
Despite the fact that much progress has been made in the research of functional customization of ultrathin two-dimensional materials for activating and converting inert small molecules, there are still many challenges: (1) How to realize the industrial-scale production of two-dimensional materials and gradually narrow the gap between laboratory research and practical applications need continuous and intensive exploration. Although a number of top-down and bottom-up methods have been employed for the synthesis of ultrathin two-dimensional materials, it is still challenging to prepare two-dimensional materials on a large scale. The mass production of ultrathin two-dimensional materials with specific surface defects will be the pivotal issue toward the photocatalytic application. More diverse and abundant synthetic strategies should be explored to prepare defect-rich two-dimensional materials with the ultrathin thickness on a large scale.
(2) Further synthesizing two-dimensional materials with improved photocatalytic activity is still urgent. Firstly, constructing two-dimensional semiconductor heterostructures is a promising approach and several methods have been investigated, like epitaxial growth.(100) It is necessary to develop the technique of constructing high quality two-dimensional semiconductor heterostructures with different functional characteristics assembled into a single nanosheet, and the effective combination of different multifunctional materials, which shows a wide range of application prospects in solar energy conversion, such as photocatalysis. Secondly, the Z-scheme heterojunction with atomically ultrathin twodimensional materials could promote charge carrier separation and retain the oxidation and reduction ability of components, thus receiving widespread attention. (101)(102)(103) However, this strategy is mainly applied in water splitting, reduction of organics, and lithium-ion batteries. (104)(105)(106)(107) In the future, by integrating photocatalysts with the strong oxidizing capacity to promote the oxidization part and photocatalysts with the strong reducing ability to carry out the reduction part, it may further play a role in the more extensive inert small molecule activating and converting ability.  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  (4) It is uneasy to establish a clear structure-activity relationship. From the perspective of structure, the structural analysis of the active sites of photocatalysts has always been an impediment in this field, which severely limits the study of the photocatalytic mechanism and structure-activity relationship, and hinders the rational design of new highly efficient photocatalysts for inert small molecule activation and conversion. This dilemma is mainly due to the complexity of the composition and structure of the photocatalysts prepared by traditional methods. At the same time, effective and accurate structural characterization methods are urgently needed for a clearer or even quantitative structure-activity relationship in the future. Atomic-scale defects can notably alter the local vibrational responses of materials and thus their macroscopic properties. Up to now, the precise signal could be detected by using such as high-resolution electron energy-loss spectroscopy in the electron microscope. (110) Extensive ab initio calculations revealed that the measured spectroscopic signature arose from defect-induced pseudo-localized phonon modes with energies that can be directly matched to the experiments. This single-atom level sensitivity could be very helpful to functional customization of two-dimensional materials by subtle regulation, and to make proper adjustment corresponding to the detection results.  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59 (5) From the perspective of activity, proper descriptor is highly urgent during the process of deducing quantitative structure-activity relationships. Although photocatalysts may have many useful properties related to the activity for activating inert small molecules, there are usually very few available parameters used for construction quantitative structure-activity relationships in an extremely diverse variable space. Many descriptor-involved tentative fitting operations are required to cover the entire variable space, and the required fitting technique emphasizes not only derivation fitting, but also the quality of the prediction of the fitting. Although these methods have not established a primary quantitative structure-activity relationship, they are helpful to examine the large amounts of data generated in research. With the development of specific methods, their applications in the field of photocatalysis will become much more widespread.